Fortinet Inc (FTNT) 2017 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Fortinet third quarter financial analyst Q&A call.

(Operator Instructions) As a reminder, this conference is being recorded.

I would now like to introduce your host for today's conference, Ms. Kelly Blough, Vice President of Investor Relations.

Ms. Blough, the call is yours.

Thanks, Sherry.

Thank you, everybody, for dialing back in.

I would like to remind everyone that the disclosures and cautionary language that I read at the beginning of the call at 1:30 today still apply to this call.

We will be making forward-looking statements, and all of those are subject to risks and uncertainties as outlined in our presentation online.

This call is primarily -- is entirely Q&A, so I will turn the call back over to the operator to moderate that Q&A session.

(Operator Instructions)

Operator, you can go ahead and poll for questions.

(Operator Instructions) Our first question comes from Jayson Noland with Baird.

Okay, great.

I wanted to follow up on the fabric.

It's come up a lot on the call today, and it's a term that's used widely across the industry.

So I guess, my question is how do you differentiate relative to others?

Is there a good way to do that in a proof of concept?

Or what's the best approach there?

Yes, this is Ken.

We have a more broad product covering different application than other vendors.

That's where we're also working on this for long time, in the last 17 years since the company started, like whether it's a mail product, FortiWeb product, the [cloud] product, so in-house developed.

All these products actually, when we develop the product, we also intend to make it working together with the FortiGate, which is a firewall -- (inaudible) firewall UTM product.

So all these kind of integrate well, and also they can also ease of the management.

So we have the FortiSIEM, also one of the fast-growing in the last couple of quarter, which also help in manage all this together.

So that's different than some other company.

They may have 1 or 2 products that are working together.

We have much broader.

Not only that, but also the access product, WiFi, like the networking part, like the cloud side and also the endpoint and the IoT side.

So all these, tied together, is the internal-developed -- intended to make it working together from the very beginning.

So that's different than some other company, whether by acquisition or by some like closely integrated product.

So I think that's where we feel the fabric start that we have is much better and work more tightly, integrate better than other vendors.

Maybe just a quick follow-up there, Ken.

Is the pricing offer concessions for bundles or volume deals, be it an enterprise agreement or something similar?

Right now, the bundle more tied to different function.

We -- yes, the enterprise, for you to have -- to say some enterprise, they may even have like a (inaudible) agreement, that's where we all -- we are open for that.

And we're definitely working together, try to see some -- may sell all this together.

But right now, it's still the early stage.

It's more like the FortiGate is still leading a lot of deals, so we land this FortiGate first, and then eventually expand into some other products, up-sell, cross-sell, but going forward could be all bundled together.

Our next question comes from Anne Meisner with Susquehanna Financial.

Great.

So I wanted to follow up a little bit on the product revenue growth.

The Q4 guidance implied somewhat flattish growth.

And if I take this year's sort of sequential growth rate and I apply them to 2018, I basically also get flat growth.

Now I know you're not guiding for '18, so I'm not going to specifically ask you, but I was hoping you could just tell us kind of how you're thinking about the product revenue line longer term.

Because if you're running the business to maximize subscription revenue, which I would argue is probably the right thing to do anyway since it's profitable, it's recurring, not so volatile in the quarter, I guess, you could ask yourself if these products really need to grow longer term.

Can it just stay flat, and then we'll let services -- the services revenue line kind of pick up the slack in terms of driving the growth for the company?

Is that -- I mean, is that one way to think about it?

Because it wouldn't necessarily be a terrible thing if that sort of what happens as long as the growth in services is more sustainable.

But I guess the question is, if that's kind of the case and that's how you're thinking about it, maybe it's better to just sort of message that a little better to The Street so it's better understood.

Fair enough, Anne.

I think what you're seeing, however, is -- it is what you said.

We're definitely seeing a mix shift as we do more enterprise business, as customers buy more of the non-FortiGate, they buy higher products, more content-rich bundles, and they buy higher level of service.

And I think I'd mentioned a couple of times now that they're buying more 7 by 24 than 5 by 8 for instance.

So what happens along with -- as you go into the enterprise, we tend to get customers buying 3-year deals or longer.

It's -- they're really just spending out because they don't want to be doing the renewals.

We're seeing that on a competitive basis, that competitive landscape.

I think our competitors see the same thing.

And what happens when you mix those 2 things together, you end up -- they end up compounding the deferral of revenue from the product line to deferred revenue.

And that's why you see product revenue at 7%, but deferred revenue, you see growing 30% and off of maybe a 1-month increase in average duration.

And so basically, that's what's happening there on the product line.

And I think that becomes harder to forecast, if you will.

But at the end of the day, it has the effect of what you're saying, where you're just ending up basically becoming more subscription-oriented, more ratable, more predictable on the revenue line, so to speak.

Yes.

Like, Anne, we are not intend to slow down the product growth.

The way we calculate maybe have a few point -- have few percentage impact.

But also, we are a little bit behind on hiring the sales capacity.

So we already realized that, a couple of months ago starting to accelerate hiring but also that takes some time to see the result.

So that's where we are a little bit crucial about Q4.

But we do intend to keeping growing the product side together with service.

Okay, great.

And then quick follow-up on the service provider vertical that's been a topic of conversation for you guys.

Some of your competitors have also discussed kind of making a bigger push into that market.

And again, I'm talking about the traditional service provider markets and not the cloud providers.

Maybe you can just kind of talk about what you're seeing in terms of the competition.

Are you seeing anything different there?

Are sales cycles lengthening because customers are just taking more time to evaluate other products in the market as well?

Maybe you can kind of speak to what you're seeing competitively specifically in service provider.

Yes.

Service provider, like a few years ago, is the biggest sector for us, like more than like 25% of the business.

But in the last couple of years, there's some changing shift in there.

The cloud provider starting with more influence in the -- compared with the traditional service provider.

So we realized this 1 to 2 years ago, starting build a team, the cloud team and service product team together and grow both.

So if you combine these 2 together, we have our healthy growth.

But the traditional service provider, they also try to position themselves compared to the new cloud provider.

That's where we feel, long term-wise, the service provider, that means together with cloud provider, is still big potential, big growth.

But somehow, the industry has some change in transition going on right now.

The traditional service provider kind of, whether they slow down some investment infrastructure, whether with 5G, and then the new cloud provider have some impact there and some of the traditional service provider also move to the cloud.

So that's the part we feel is not quite settled down yet.

And we see a lot of new testing going on, but they are not quite reached some of the decision of like a big deployment yet.

Our next question comes from Melissa Franchi with Morgan Stanley.

Just a quick follow-up on the discussion on the product revenue and the deferral.

So when you do have to defer a higher percentage of product revenue than you've seen in the past, I just want to clarify, does that go into product deferred, and then eventually it gets recognized into product?

Or is that really the percent is moving to subscription, the value goes into subscription, and then that eventually flows through the subscription revenue line?

It is -- good question, Melissa.

And just to back up.

I mean, it's a -- by the way, all of this is a characterization of an improving enterprise business, to be frank, right?

We're getting more complicated deals, they're buying more -- and they're buying more of the services and subscriptions.

To answer your question directly, yes, it goes to -- it ends up being in services and subscriptions.

So it gets, I'll use the word carved out, a product, and then recharacterized as services and subscriptions and recognized over time, not back into product.

Our next question comes from Michael Turits with Raymond James.

A couple of questions.

First, I just want to -- just a -- kind of a housekeeping on CapEx.

I just wondered if we're still in line for doing $150 million or so this year, maybe $90 million next year and then growing from there.

We're -- I didn't -- Michael, we're in line for $150 million this year.

I didn't hear the second part of the question.

And then I have it at $90 million for next year, and then growing...

Yes, we haven't guided for next year, but I think we've said -- I suspect the way you're getting there, and not to put words into your mouth, but it's probably roughly $30 million of general CapEx and maybe $60 million of real estate.

Because I think we said $50 million and $60 million of real estate for that next couple of years, if I go back to the last calls.

Okay.

But then it's...

That would get you to about $90 million.

Again, we haven't formally guided for next year, but I think people ask -- tend to ask us what's our run rate CapEx, and it generally weights out to be around $30 million, could be plus or minus a little bit, and then the real estate has -- as we build out the headquarter property, we've said $60 million and $60 million for, I think, '18 and '19.

Okay.

And then I just wanted to understand, Ken, you're talking about the traditional service providers, I assume you mean carriers, and then you were talking about the cloud service providers.

But I didn't understand, was your point that the cloud -- that you're starting to sell more to the cloud service providers as opposed to the carriers, and that -- but that, that's not making up the difference?

I was trying to -- I didn't understand what you were trying to convey.

We category a little bit differently the cloud provider and also the traditional service better.

I think, so far, the service provider we're still more common the traditional service provider.

Not common cloud yet, right?

The service provider line, the vertical.

Yes.

So that's where the one we gave out, I think it was 17%, just the traditional service provider.

There's no cloud provider there.

But they do have a cloud provider starting to -- some of the...

It's actually in there.

I think it includes -- from a provider perspective, it's in there, right?

If it were a virtual license, it would either be a product or on the subscription services line if it's ratable.

Okay.

Yes, we can confirm with you later.

Yes.

Yes, yes.

Okay.

And then another question.

So it sounds as if one of the major factors is lumpiness in the enterprise business.

Longer sales cycles, bigger deals, more complex.

And this sounds like something that you went through a couple of years ago, where enterprise got really good, and then it got tougher.

And the way you explained it at that time was that you were doing well with getting -- winning a lot of reserve -- large, high-end box and larger deals.

But you weren't getting, I think you termed it as the -- one of the run-rate enterprise business that would drive more consistency.

I'm wondering if is that the same dynamic that we're seeing this time around?

I think it's different.

I -- they're -- what we're seeing now is more complexity because they're thinking architecturally.

Again, the -- to step back and you think about security spend, I think Ken mentioned earlier that, post-Target, everyone -- there was a lot of reactionary buying.

What you're seeing now is more strategic buying.

And the fabric resonates extremely well because, as Ken said, it's broad, integrated, and an open architecture, if you will, provides the management capabilities.

And so if they buy the full package, the fabric or more of the fabric like the example I gave, they end up with more products.

And then from a forecasting perspective, I think it becomes more of a challenge because they may buy some or all at the same time.

And I think that's where it generally, from a sales perspective, becomes harder to forecast.

They believe they may get a customer for instance, but forecasting the actual deal size when there are a variety of products on it may shift some in or out of a quarter, and like you said, some of it may become more run rate, some of it may become more upfront.

But as I think architecturally, I think this just becomes more of a challenge to basically forecast at a basic level.

Those -- again -- and to revisit, again, we drilled into -- we did a very good job, I think, of drilling into our geographies and trying to understand what was going on in the business.

And I'm going to repeat myself here, but we did -- when we look at North Asia, we saw some macroeconomic issues there, really, we believe due to geopolitical issues.

And then we're talking to our people on the street there, and that's what they're telling us.

And in Latin America, we have a sizable business, especially in Mexico and somewhat in the Caribbean, and we saw some softness there in Q3.

And we're -- the team is telling us to call that into -- call that forward into Q4, and that's a bit of the explanation.

And then as Ken said, just to go back, we're a bit behind on headcount.

And I think all of those factors weigh in and blend together, and it's hard to weight one ahead of the other, but they blend to form the forecast or the guidance this quarter.

Our next question comes from Erik Suppiger with JMP.

I know you don't want to give guidance for next year, but your gross margin guidance has bumped up over the course of the -- of fiscal '17.

Is there any reason why it would reverse and go back down in '18?

Or are there any trends that were anomalistic in '17 that we should be factoring in?

And then similarly, your guidance for your -- you did about 18% to 19% operating margins for the -- for Q2 and Q3, and that's your guidance for Q4.

How can we think of trends in terms of your hiring as you look into fiscal '18?

Are you -- is it likely to come down because you're going to be hiring aggressively?

Or how should we think about that?

Why don't we -- so I think there's 2 questions in there.

There's a bit of a headcount question, I think, Erik, is the second part in that ramp, and I believe the first part is related to gross margin?

Yes.

And so on the gross margin, one of the interesting things is it actually would've been better if we didn't have the carveouts that Anne and I were discussing earlier.

Because effectively, we're recognizing the cost of goods sold, the hardware cost upfront, yet we're deferring more of the product revenue onto the balance sheet.

And so that margin ends up coming back over time in that strong, ratable revenue recognition, if you will.

So structurally, the way to think about it is, structurally, our business is driving higher margins all around with this mix shift to one more enterprise and then more fabric.

And that fabric's composed not only of non-FortiGate products, but more margin-rich services and subscriptions.

And if you blend that together, Erik, you end up with a more -- with a higher-margin business because of -- most of that's higher margin.

And to the extent that it -- that some of that's product revenue coming back on through the services and subscription line, I've already taken the cost of goods sold in the quarter I shipped it.

So that's the first part.

Headcount, we're behind.

And again, we have kind of a -- we believe, structurally, our business has changed, it's driving higher margins.

And we need to make sure that we catch up on headcount so that we could continue to grow.

And so we're balancing that out, and we'll talk about guidance when we get into '18.

Can you just comment, though, how far behind are you on headcount?

I mean, would you -- I think you added over 100 this quarter.

Would you have liked to have added 300?

Or how can we think about the plans on the headcount front?

I think to supporting the -- some of the growth target we have, we probably need to add a little over 100 a quarter.

But in the first 2 quarters, we are less than half of that.

That's also making like a margin probably above the target and the Q4 guidance will be low.

So we realize that -- we starting our salary something in Q3, but it's Q4 we need to keep in -- catch up some of the headcount hiring.

Our next question comes from Fatima Boolani with UBS.

Just a question for you on the product gross margin.

You did talk a lot about the traction that you're seeing in the enterprise, and we saw the product [skew] mix, skew very nicely to the high end of your portfolio.

But the product gross margins were essentially flat sequentially, so is there any other dynamic that we need to consider that's kind of going into that line?

And then just a quick follow-up on free cash flow and any modeling points you could share with us.

Okay.

So Fatima, I think I heard 2 questions.

Product gross margin, and then I think you're -- free cash flow guidance.

Fair enough?

Correct.

Yes.

Good.

So what I was attempting to explain to Erik earlier is that, basically, because we end up with a combination of more content-rich services being sold, if you will, more FortiCare -- higher level of FortiCare support 7 by 24 versus 5 by 8, and then richer content in the subscriptions that are higher priced like the enterprise bundle, for instance.

That combined with duration, longer duration, we shifted from 24 to 25 months, causes more of a carveout, if you will, more of a deferral from the product line being recharacterized onto the services and subscriptions line, if you will, the services line ultimately, but services and subscription.

And what that does is you ship it -- while we're effectively taking that product revenue over time on those lines, the cost is recognized upfront.

And so if you really looked at it on a deal-by-deal basis, the margins actually are better than reflected in the gross margin line.

It would be higher, for instance, if you were, if you just looked at it on a billings perspective.

And that's really what's going on there.

There were no other anomalies in the quarter.

There were no -- I think, last quarter, we referenced taking some reserves on inventory.

We didn't do anything out of the ordinary on that front this quarter.

It's just really reflection of some of that product revenue ending up in deferred revenue, which grew 30%.

Very clear.

And just any directional color you could share on free cash flow and how that shakes out?

Just because you had a very strong collection quarter this quarter.

What we guide is -- to be frank, we've been focused on linearity and cash conversion cycle, Fatima.

It becomes a little harder to predict because there's a lot of timing differences in there.

I think I could share CapEx.

We end up -- we expect still to end up at CapEx at $150 million for the year.

I think -- yes, and we've spent $122 to date, I believe.

Our next question comes from Ken Talanian with Evercore ISI.

I was curious, have you made any changes to either the frequency or the methods you're using to evaluate your pipeline?

No.

We are -- no, Ken.

We continue to look at it very closely.

No, we haven't made any changes.

Okay.

And then just as a follow-up, obviously there's a greater shift to ratable revenue.

Can you give us a sense for -- maybe for 4Q, how much of your ratable revenue is expected to be rolling off the balance sheet from deferred?

I don't think we've ever released that metric, how much of the in-quarter revenue came off the balance sheet.

I mean, it's kind of -- it's reflected in the services line.

If you look at the balance between products and services revenue in an individual quarter, it should sort of reflect that percentage.

Our next questions come from Patrick Colville with Arete Research.

I've got quite an easy layout one first.

You guys are still growing product revenue faster than the market.

Who are you taking share from at this -- right now, October 2017?

I think, if you look in the first half of this year, our billing growth probably faster than any other major security vendor.

And so probably we take some share.

I think anyone grow below the market rate, we take share from them.

I cannot name the...

I would say we're playing on the theme of where integration and that organic integration really sounds well, where we can get the visibility and -- Patrick, and that tends to be large enterprises.

And so that could be a variety of incumbents, quite frankly.

And I think you could probably guess the names.

Yes.

The enterprise section grows faster than the other section, so that's where some traditional enterprise player we take quite some share from there.

Got it.

Okay, sure.

And then if I can do a quick follow-up.

And also, you start in U.S. I have to say, you can see U.S. is a fast-growing region, so that's where we take most share from some of the U.S. player.

Got it.

Okay.

I mean, I guess, the basis of the question is that some investors are nervous that Cisco and Juniper have less share to give up, and one of the reasons for kind of the slowdown in the product growth is because there's less share to take off those guys.

Yes.

Some of other competitor also, their U.S. slowdown probably below the market growth.

We also take some share from there.

Okay.

And then a true quick question on the share repurchases.

It looks like you guys are doing that more consistently now, and the kind of dollar amount increased this quarter, for sure.

And is that something we should kind of think about modeling forward, just much more consistent quarter-on-quarter share repurchasing?

Well, as we said on the call, the board, I think, is signaling a more aggressive approach to share repurchases.

And we -- they approved up to $1 billion, the current authorization.

I think we have $765 million relating -- or had as of the end of September.

And so as far as consistent, I think it's a little harder to say but more aggressive for sure.

Our next question comes from Sterling Auty with JPMorgan.

This is Ugam Kamat on for Sterling Auty.

I wanted to understand the price difference that happens when you actually sell a virtual firewall in lieu of physical firewalls.

Assume that you're going into an enterprise and selling a physical box, and in lieu of that, if you sell a virtual firewall to them, what is the pricing economics on that for the same throughput?

Well, Ugam, the way we answer the question is that it's an accretive model to the overall business.

Because right now, what we're selling -- we're seeing multi-cloud environments where they're buying a variety of products.

And the -- whether it's a virtual -- the cloud orientation is pulling the hardware as well because it's a hybrid model.

And so the economics work out that it's accretive to our model.

The other way to think about it is the -- to the extent that there's virtual, it's a higher-margin business.

Okay.

Okay, that's pretty helpful.

And secondly, just to expound on the share repurchase program, in terms of assuming, are you planning to do share repurchases that offsets the dilution in share count or are you willing to go beyond that and do more share repurchase?

Like to offset more than the dilution?

Right now we are focused on being more aggressive on the share repurchases.

Again, I'll repeat what I just said and I'll have to leave it at that.

The -- and you could see, first of all, the board approved up to $1 billion of share repurchases, we have $765 million remaining.

You could see that we're becoming more profitable on a GAAP basis, and we are very focused on the equity comp aspect of the dilution, if you will.

And that's how I would answer that question.

Our next question comes from Howard Smith with First Analysis.

Yes, I wanted to follow up on kind of the other side of the gross margin, not product but services.

That continues to grow, and I know part of that is the opposite of what we just talked about, the deferred coming through and being very high margin.

I calculated about 95%-plus incrementally this quarter on a non-GAAP basis.

Is there some headcount adjustment, a catch-up that needs to be done that would affect that line?

Or is it just that profitable as the incremental services come in?

Well, we're effectively -- I think you're getting scale benefits is the way to think about the answer, because you're really getting more revenue on a not completely constant headcount or infrastructure basis.

There are some scale investments, there are some investments you need to make in headcount every time.

We're probably a little behind on that, which helps a little bit, but I don't think it's significant and probably not the right way to think about it.

But a little bit behind on headcount on the support side, if you will.

We've become more efficient all around on just managing the business.

And things like RMAs, for instance.

Our RMA costs are, I think, managed better over time and so we're driving efficiencies like that.

And so even though you might hire more people, and it's not a huge ramp there, we're finding ways to offset it by driving other efficiencies in the business.

And then you offset it again by just increasing revenue on the services line, what we talked about, that mix-shift benefit.

Our next question comes from Saket Kalia with Barclays Capital.

So first through, just on the billings guidance.

I've just got -- sort of got a philosophical kind of question on the approach.

If you compared your approach to billings guidance today compared to the approach that you took at the end of Q2, right, we raised it in Q2, we're trimming it now.

If we exclude some of the regional stuff, has the pipeline for enterprise this year gone down?

Or is it maybe more prudent close rates on that enterprise pipeline than you expected before?

Does that make sense?

Yes.

So if the pipeline goes down or do you need higher -- I think are you asking if you need higher close rates?

Right.

Or are you assuming lower close rates in the fourth quarter guide than maybe you did before?

Not really.

I think, overall, we're going off -- it's a variety of things.

I mean, that's one of the factors in there, Saket.

But I -- we're really relying on our sales force to tell us what they're going to do and call the shots.

And I -- when you dig into it, as you said, take out the macroeconomic issues by the various geographies and the carrier business, what there's -- they're all trying to sell the broader fabric.

And I think it just becomes more complicated on exactly what a company is going to buy at any given point in time.

And so that really creates the complexity we have in forecasting going forward.

I would point to that as the issue.

Got it, got it.

And then a quick follow-up, if I could squeeze it in.

And this is -- forgive me, it's probably a dumb question, but I just wanted to understand the pricing just a little bit better.

I guess, on the carveout accounting, I think we get sort of the idea of additional content that enterprise customers are getting.

You mentioned 7 by 24 instead of 5 by 8. But if an enterprise customer is getting more, wouldn't the carveout be of a higher number?

Or are we maybe discounting that so the carveout is actually impacting the same number?

Could you just talk about that number -- that approach to pricing a little bit?

Maybe that could help clear some of that confusion about sort of the carveout in the product.

Sure.

Saket, I'm going to ask you to -- if you could -- I didn't quite -- I'm trying to answer the right question.

If you could perhaps maybe ask it again?

Yes, sure.

So can you talk a little bit about the pricing in a carveout situation?

So in a situation where an enterprise customer maybe opts for more content-rich items, like an enterprise bundle or 7 by 24 support, that is -- that would presumably be additive to a deal.

And so the carveout, right, would be a carveout of a larger number.

But I guess, it almost implies that maybe we're seeing larger carveouts of the same number.

Am I thinking about that right?

I think 2 factors.

I would say 2 factors.

Generally speaking, if we're going to discount on a deal, and we could talk about discounting if you want, it's not a big issue but to the extent there's always some discounting, the discount would tend to always -- tend to be on the product line to preserve the (inaudible) on the recurring revenue lines, which is the subscriptions and services.

And so -- or you do the -- excuse me, as you discount across the board, you end up carving more out from product to deferred.

And I think the compounding factor tends to be the duration that we pointed out, quite frankly.

And that's what's going on where the duration impact -- we've gone from 24 to 25 months, I think, average duration.

That compounds.

So on a 3-year deal, you end up carving out 3x of whatever you were going to carve out and regardless of where the discounting occurred.

Our next question comes from Gabriela Borges with Goldman Sachs.

A quick follow-up on Saket's question, actually.

So for Drew, one of the things that has been mentioned a couple of times is that enterprise deals are becoming more complex, they're a larger piece of the business.

Given that, that seems a little more structural or ongoing in nature rather than temporary, I'm wondering if there are business intelligence tools or forecasting tools that could maybe help internally to drive a little more consistency in forecasting, given that it feels like this is something that will be ongoing over the medium term.

I mean, I think it's fair to say that we kind of -- we know what tools are available at our disposal.

We do pay a lot of attention to forecasting.

We have regular meetings with our executive team and our sales team to go over the forecasts.

I mean, it's a -- we get your point, Gabriela.

In this particular case, it feels more like there were things that weren't in the forecast that came out because of, for example, the things that Drew mentioned, the Mexico environment, Latin America and other parts of the world.

We take your input, Gabriela.

I would say this, we're very diligent in drilling down on the forecast and getting the explanation from our team and our sales team and understand what's going on all around.

And at the end of the day, that's the most important thing is really getting the story and getting what their commitments are and why.

And effectively, that's the story we're relaying to you.

Our next question comes from Catharine Trebnick with Dougherty.

I apologize if it's been asked, we've had 5 calls tonight.

One thing I noticed during the quarter was channel, there was a -- you brought back a person that had been at Fortinet then promoted him.

Can you just give a little -- an update and more color on the channel strategy?

Is it pretty -- where are you headed with it, U.S. versus Asia versus other geographies?

Yes.

John Bove came back.

He was excellent.

He came back from Proofpoint and an excellent employee.

Nothing's really changed in the strategy.

We believe our business is doing fine there.

We look at that in detail.

We haven't seen any issues there, and we believe John's a [great] addition, coming back into the team, and we think we're fine.

Is there a second part to that, Catharine?

No, I was -- that was really what I was looking for.

I noticed he came back and, obviously, he likes working for you, so congratulations.

Our next question comes from Hendi Susanto with Gabelli and Co.

Drew, product gross margin is lower this year than last year, and you alluded that it's driven by the shift of products to deferred revenue.

Should we see this gross margin trend to amplify to a larger degree down the road?

In other words, whether product gross margin may slide down further as this trend grows?

The way I would think about it is to recharacterize revenue that, in some way, causes a slightly lower gross margin this quarter will shift the trend -- that's on the product line -- slightly lower gross margin on the product line this quarter, excuse me, will tend to come back as ratable revenue over time.

And really, I think you want to look at them in tandem as a blend of gross margin.

Because it comes back as a higher gross margin altogether.

And if you look back over the last several years, gross margin's gone up from Q3 of '14, about 71.5%, to 76% this quarter.

And so that's a bit of what's going on.

And if I look back over the quarters, 59% is not out of line with some of the past quarters.

Got it.

And one question for Ken.

With regard to the challenges of less predictability and more complexity in enterprise deals, do you see this as unique to Fortinet?

Or do you see this as a general situation that applies to your market and your competitors?

We just need to get the sales force more trained for some enterprise, especially the new sales force.

We're starting our [salary] hiring, so that's taking a little bit more time to ramp up.

That's where -- like I said, you can't improve margin quickly, easily probably the same quarter or the next.

And -- but we invest in growth, that's take about 6 to 12 months.

And so if we are behind early at the year, we're catching up, and that's probably we'll not see the result in the quarter right away.

And then if I understand that correctly, so that implies that partially it's due to the sales force situation, and partially it may be driven by the customers in the market, overall?

I don't mean to put your -- some words on your mouth, but is that the right interpretation?

Yes.

Also, sometimes, we sell multiple products.

It's a little bit more complicated and maybe take a little bit more training and more time compared to traditionally more single firewall product.

Our next question comes from Andrew Nowinski with Piper Jaffray.

All right.

Your operating cash flow, you said, is going to grow in line with the operating income going forward.

But obviously, in 2017, it looks like operating cash flow clearly outpaced operating income.

Do you think it'll more closely track operating income growth in 2018 and beyond?

Well, when you step back, I think what -- a couple of things have happened.

One, you -- as you said, I think you've had some margin expansion, if you will.

But -- and that characterized part of it.

And we've also done a really good job, I think, on cash conversion cycle, growing -- basically increasing our working capital.

We'll continue to focus on that, Andrew, but it's a little early to really call that for next year.

All right.

Fair enough.

And then DSOs continue to decline as well despite the lumpiness in the large deals that you called out.

So what are your expectations for DSOs going forward?

Well, again, we've been focused on, as I would say, cash conversion cycle.

And that includes linearity as a component and working with sales to drive more even linearity, and we saw that in Q3.

It's a little harder to predict what'll happen in Q4.

Quite frankly, it'd be nice to see that continue, but it'd be a little early to call that one, too.

But we remain focused on it, to be fair.

We do.

We talk about it.

All right.

Then, Ken, I think you mentioned that you start to see a refresh cycle having a more material impact 2 years out.

Are you seeing any customers thinking about upgrading their firewalls to get better SSL decryption performance, given the increasing amount of encrypted traffic that's coming into a data center now?

Yes.

Some service provider, some cloud provider, for them actually the SSL encryption acceleration is very important.

And you using the software approach, difficult to achieve that high performance.

So with our SPU, the security processor, they can run that acceleration 10 to 100x faster and probably -- and even at a lower cost after 10 to 100x acceleration.

So that's where we see quite some opportunity and also business from there.

So it's SSL definitely driving some of the major growth, especially on the e-commerce and the web-based.

Our next question comes from Erik Suppiger with JMP.

My questions have been asked.

Our next question comes from Fatima Boolani with UBS.

Drew, just a question for you on some of the longer-term targets that you've talked about historically.

So you're tracking at about [18%] billings growth this year, but the longer-term aspiration is to hit kind of the 20% mark.

So the acceleration there, can you help me understand what the input would be to help you drive toward that?

And maybe contrast that with some of your comment around you being behind on hiring as well as some of your comments around the 150 to 200 basis points of [brand] expansion on the operating margin.

So just help triangulate your billings acceleration, you being behind on hiring, but then also sort of driving operational efficiencies to drive margin expansion.

Sure, Fatima.

The -- structurally, our business is driving higher margins, and you could see that primarily from the gross margin impact.

And I think we've spent a lot of time talking on the mix shift and more enterprise business and how that reflects more on the deferred revenue, where you have some of the upfront margin due to the product side over time, and then you take the ratable revenue comes back really into your margin over time.

And so that's a tailwind to the business, and it's a structural change that's driving.

Also, you get more subscription and services revenue.

You're not on a completely flat cost basis on the gross margin line, but it drives -- there's scale benefits in that business.

And you've seen the leverage as we move forward on the operating lines as well.

What -- in terms of headcount, we absolutely need to catch up, and that would really be the question in the balance at this point in time.

Got it.

And if I could sneak another one in.

Sure.

You talked at a high level about the penetration success you've seen with the enterprise bundles as well as the SMB bundles.

You have a tremendously large base, and I'm -- and historically, because you sold your security functionalities sort of a la carte, if you will, can you help us understand what proportion of the base has still not had a chance to buy into the bundle proposition, whether it's SMB or enterprise?

Yes.

There's actually a couple of things, Fatima.

One is the enterprise bundle.

Then there's also selling a higher level of support, if you will.

We talked about 7 by 24 versus 8 by 5. And then we continue to focus on driving -- developing more higher-priced, higher-content bundles, if you will, which seem to be resonating very well as we continue to sell the fabric.

And so the -- I would characterize it as early, still in that process.

We believe we have a way to go to up-sell and cross-sell into -- that into our installed base.

We're very focused on it, and so there's plenty of opportunity there.

Our next question comes from Ken Talanian with Evercore ISI.

Just a quick question.

Have your hiring plans been altered at all by your real estate investment plans?

I mean, is there a correlation there?

No.

Our next question comes from Anne Meisner with Susquehanna Financial.

Drew, quick follow-up for you.

I don't think we've discussed everyone's favorite subject yet, which is ASC 606.

And I know that the last time you gave an update, you felt that there wasn't going to be a ton of change in sort of the recognition part of that on the rev rec side, I guess.

But since we're kind of talking about products recognition and carveouts, I just wanted to ask you again.

Because some other security vendors have talked about how 606 may require them to ratably recognize the appliances because they would be kind of tied to the subscription components, and I'm just wondering if that's something that might end up being the case for your model as well.

We're not finding that to be the case, Anne.

Obviously, we're not through it, but right now, we would still say that we don't see a significant impact on the revenue line.

Thank you.

I'm showing no further questions at this time.

I would now like to turn the call back over to management for any closing remarks.

Thank you again, everyone, for being on both calls today.

And if you have any follow-up questions, please feel free to call me.

Thank you.

Bye-bye.

Ladies and gentlemen, thank you for participating in today's conference.

This does conclude the program.

You may all disconnect, and have a wonderful day.